Assessing the Relationship Between Propagule Pressure and Invasion Risk in Ballast Water (2011)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

5  Other Approaches to Setting a Ballast Water    Discharge Standard  The EPA and USCG seek a scientifically based process for determining a ballast water discharge standard that will protect against the establishment of new nonindigenous aquatic species in the nation’s waterbodies and protect and preserve resident species and other beneficial uses in these systems. In the ab- sence of data and models necessary to support a science-based quantitative ap- proach (see Chapter 4), expert opinion has been a common alternative to model- ing to advance discharge standards. Expert opinion can be defined as a process by which a person, or team of individuals, with a high degree of knowledge pertaining to a particular topic draws upon that knowledge to propose an estimate. In the area of standards for living organisms in ballast water discharge, expert opinion processes have taken a number of forms and, as observed by Lee et al. (2010), produced a wide range of outcomes. This chapter discusses strengths and weaknesses of the expert- opinion approach and reviews three expert opinion-based standards for living organisms in ballast discharge: the (1) IMO standard setting approach, (2) Cali- fornia’s Zero Detectable Living Organism standard setting approach, and (3) Natural Invasion Rates as a basis for a standard. The Natural Invasion Rate me- thod put forward by Cohen (2005, 2010) starts with a desired invasion rate based on expert opinion, and then combines it with a quantitative risk–release relation- ship (much like those discussed in Chapter 4), to derive allowable discharge levels. This chapter considers only the expert opinion-based portion of the me- thod. EXPERT OPINION AS AN APPROACH TO DECISION-MAKING Expert opinion has been applied in the field of biological invasions in a number of ways. First, expert opinion has been used to identify nonindigenous species that may pose problems if introduced to a new country (see IUCN, 2000; European Environment Agency, 2010; and Kolar and Lodge, 2002). The Cana- dian government developed a risk assessment for silver and bighead carp species   114  

Other Approaches to Setting a Ballast Water Discharge Standard   115    based upon expert opinion, which resulted in a regulation banning live sales by fish retailers (Mandrak and Cudmore, 2004). Expert opinion also has been used to screen for invasive, weedy plants and insect pests, among others (see Tucker and Richardson, 1995; Daehler et al., 2004; Paini et al., 2010). An inherent advantage of the expert opinion approach is that it can produce estimates in the absence of robust data sets. Some models based on expert opi- nion can be cross-checked using other methodologies, allowing researchers and managers to consider the biological realism of their predictions (Phelong et al., 1999; Paini et al., 2010). Nonetheless, there often remains considerable uncer- tainty in the output from expert opinion, especially when there are critical gaps in understanding underlying mechanisms. Expert opinion can be greatly affected by problem framing, context depen- dence, and motivational bias by participants (Burgman, 2005). Thus, a key problem is that very different outcomes may result from different parties consi- dering the same problem. It is also difficult to test the output for validity since critical considerations are often implicit in the process. Consensus-based expert opinion by cross-disciplinary teams may help eliminate blind spots in the analy- sis in terms of types of expertise, but these exercises often must limit the number of experts per discipline to one individual. As a result, inherent biases within an expert become a greater concern. Maguire (2004) concluded that risk assessment based upon expert opinion is often conflated with risk management, and that assessments of risk based on individual judgment often incorporate personal outcome preferences. A further problem is the influence of incomplete information on decision making. For example, highly invasive species including the emerald ash borer (Agrilus pla- nipennis) and quagga mussel (Dreissena rostriformis bugensis) had little or no invasion history prior to colonizing North America, and expert opinion based on invasion history likely would have concluded that these species pose little risk. While risk assessments based upon performance (effect) in native regions may be informative and are commonly conducted, the colonization process can result in evolutionary (genetic) changes, where native and introduced populations dif- fer substantially. Moreover, invasions by definition place species in novel envi- ronments, with a different assemblage of organisms. Thus, native populations may be poor predictors of population dynamics, interactions, and effects in an invaded area. Data on the concentration of living organisms in ballast water and on the probability of invasions occurring in U.S. waterbodies as a result of ballast water are extremely limited. Thus, despite its drawbacks, expert opinion has unders- tandably played a major role in ballast discharge standard-setting to date, and it will likely be involved as part of the risk management decision-making process into the future. 

116  Propagule Pressure and Invasion Risk in Ballast Water    IMO STANDARD SETTING APPROACH The best-known example of an expert opinion-based standard setting process for ballast water is the agreement reached at the 2004 Diplomatic Confe- rence of the International Maritime Organization (IMO), which included 74 IMO member countries, the European Community, and 18 non-governmental organizations (Gollasch, 2005). This agreement was formalized as the IMO International Convention for the Control and Management of Ships’ Ballast Water and Sediments and was signed by 67 IMO member countries. As of May 2011, the Convention has been ratified by 28 countries collectively accounting for over 28 percent of global merchant shipping (IMO, 2011). The IMO Con- vention will come into force one year after not less than 30 States representing 35 percent of the world’s merchant tonnage have ratified without reservation. In this rendition, the expert-opinion process involved at least two sets of ex- perts (there was some overlap) and two sets of rules. Specifically, the IMO Ma- rine Environment Protection Committee (MEPC) requested the assistance of a Study Group (on Ballast Water and Other Ship Vectors) of the International Convention for the Exploration of the Sea (ICES), and also a group from the shipping industry to propose a starting point for IMO negotiations to achieve discharge standards. The ICES Study Group, predominantly aquatic invasion biologists, reviewed available data sets to date related to concentrations of or- ganisms in untreated ballast, characterizing the range and distribution across many vessels for different types of organisms. This expert group recommended that to “significantly reduce the risk of invasions associated with ballast water beyond the present situation, permissible discharge concentrations identified by any treatment/management standards should fall greatly below the median val- ues observed presently for untreated/unmanaged ballast water.” This would have the effect of reducing high concentration discharges, and the group consi- dered empirical and theoretical evidence indicating that invasion likelihood in- creases with increasing concentration (International Council for the Exploration of the Sea, 2003). Recognizing that biological invasions via ballast water occur across a broad span of phyla and size classes, and that organisms comprising these groups dif- fer tremendously in abundance in both nature and ballast water, the ICES study group recommended that the IMO performance standard classify biota contained in ballast water into categories based on size or taxonomic group. This group made recommendations of discharge standards for two of these sizes classes: greater than 50 m (zooplankton) and between ten and 50 m (phytoplankton) to achieve concentrations three orders of magnitude below observed median values in untreated ballast (the recommended standards correspond to 0.4 zoop- lankton per m3 and 13.3 phytoplankton per liter). Because the recommended approach considers total abundance of organisms in a particular size class (i.e., a guild) in a ship’s discharge, the risk associated with release of any one species would be expected to be lower since ballast water communities usually comprise numerous species. 

Other Approaches to Setting a Ballast Water Discharge Standard   117    The first stage of the process for crafting the IMO standard represented the consensus expert opinion of the ICES Study Group experts involved. The out- put of this analysis was then transmitted to the relevant IMO committees and working groups to be considered along with operational concerns cited by indus- try. The standard that emerged from this two stage process places numerical limits on abundances in treated ballast effluent of less than ten viable organisms per m3 greater than or equal to 50 μm in minimum dimension (zooplankton and zoobenthos) and less than ten viable organisms per ml less than 50 μm but greater than or equal to 10 μm in minimum dimension (including phytoplankton and protists). The IMO also incorporated into the standard toxigenic Vibrio cho- lerae (serotypes O1 and O139), the etiologic agents of pandemic cholera, and two routinely used fecal-indicator bacteria, E. coli and intestinal Enterococci. These latter two are frequently used to assess the safety of recreational surface waters, e.g., swimming beaches. Some are concerned that the allowable levels of live organisms in the standard generally are too high to be protective. In par- ticular, the USCG during negotiations for the IMO agreement indicated that standards needed to be considerably more stringent (i.e., 100 times) to be protec- tive, approaching more closely recommendations from the ICES Study Group. Regardless, it appears that the IMO D-2 standard represents a significant reduc- tion in concentrations beyond ballast water exchange, especially for the largest size range (International Council for the Exploration of the Sea 2003; Minton et al., 2005). ZERO-DETECTABLE DISCHARGE STANDARD The State of California’s existing and anticipated standard for larger plank- tonic organisms in discharged ballast water is “zero detectable” living or cultur- able organisms. Like the IMO standard setting process, the process associated with the genesis of this standard also involved the opinions of technical experts (biologists) as well as industry and other groups (California State Lands Com- mission, 2010). The expert group recognized that (a) the probability of invasion declined with concentration and (b) there was great uncertainty surrounding the risk–release relationship. This group took a precautionary approach, while ac- knowledging that enforcement against a standard of zero living organisms has practical constraints. Unlike the IMO D-2 standard, the zero detectable dis- charge standard applies to all size classes and taxa (if the word “living” may be considered applicable to viruses) and is to be put in force by January 2020 (Faulkner, 2010). [The interim standard is zero detectable living organisms for taxa >50m, 0.01 organisms per ml for sizes between 10 and 50 m, and no more than 103 bacteria and no more than 104 viruses per 100 ml of ballast water discharge. Interim standards are to be in place between 2009 and 2016 (Faulk- ner, 2010).] This approach to a standard is appealing in its straightforwardness, but as Lee et al. (2010) argue, its rationale is predicated on an actual zero discharge, 

118  Propagule Pressure and Invasion Risk in Ballast Water    while its practical application can differ substantially from zero detectable dis- charge. The zero-detectable concept is constrained heavily by the sampling re- gime employed and associated sensitivity, which will depend on the analysts’ technical capabilities and financial resources. Recent advances in molecular ecology, such as use of species’ environmental DNA (e.g., Mahon et al., 2011; Jerde et al., 2011), are expanding our capability to detect rare organisms in wa- ter, although none can yet unequivocally decipher whether all organisms de- tected are living or dead. Because of the practical constraints in detecting living organisms when they are rare, the zero-detectable standard may be operationally no more stringent than other standards. The lower the density of living organisms achieved in the effluent, the more difficult the task of detecting organisms and determining whether the water is compliant with the zero-detectable standard. Lee et al. (2010) nicely summarized this by estimating the uncertainty inherent in sam- pling different volumes of water (their Table 18). For example, even should a 10 m3 sample (a huge sample) of ballast water contain no organisms, the actual concentration of organisms in the tank could be as high as 0.3 per m3. Arguably, there could be fine tuning of the Lee et al. (2010) analysis should the distribu- tion of organisms be found to be other than Poisson, but any such change would only alter the levels of uncertainly associated with a particular sampling regime. More recently, Miller et al. (2011) also considered organisms with a Poisson distribution in ballast water effluent. They demonstrated that the density of or- ganisms and volume sampled had strong effects on decisions to reject whether ballast water is compliant with a selected standard, while the permissible type I error rate (i.e., false positives or rejection of compliant ballast) had slightly less influence as modeled. Thus, while a laudable ideal, the zero-detectable standard is functionally de- fined by the ability to characterize concentrations of organisms at low densities. If organisms are detected, it is clear a ship’s ballast effluent is in violation. However, critically, the absence of live organisms in a sample or set of samples does not provide sufficient information to accurately assess densities, and there is always a non-zero probability that organisms are present below some thre- shold (defined by the specific sampling effort imposed). Thus, while Califor- nia’s standards clearly represent a reduction in concentrations below those ob- served in ballast water exchange, the exact discharge standard is largely unde- fined and contingent on sampling protocols, representing an operational defini- tion that is driven by sampling statistics. NATURAL INVASION RATES The Natural Invasion Rate (NIR) approach to standard setting incorporates both a risk assessment and a risk management component. The risk assessment component assumes that an empirical, linear relationship can be determined be- tween the rate of organism release in ballast water and the resulting rate of new 

Other Approaches to Setting a Ballast Water Discharge Standard   119    species invasions (see Chapter 4). The risk management component proposes an acceptable target invasion rate based on expert opinion from which the corres- ponding target discharge rate is determined. In this case, expert opinion has concluded that the natural (i.e., non-anthropogenic, background) invasion rate should be the target ballast-mediated invasion rate. This discussion focuses on this latter component of the approach (i.e., the risk management step). Cohen (2005) identified some difficulties inherent in estimating natural in- vasion rates and provided coarse estimates based on expert opinion that span three orders of magnitude. Providing any validation or better assessment is problematic because there is insufficient information to determine a natural in- vasion rate. The scale of available paleontological data is inconsistent with that needed for application of a NIR standard. For example, natural port-to-port in- vasion rate data are required, whereas available data refer to large-scale disper- sal events (e.g., trans-Pacific regional migrations; see Vermeij, 1991) and, thus, are not an appropriate comparator. It is expected that natural estuary-to-estuary trans-Pacific migrations would be much lower than the measured regional mi- grations. While some natural invasion rate data are available, in terms of Pleis- tocene migrations and faunal shifts along the California coast, including ancient embayments (Roy et al., 1995), they do not include trans-Pacific species move- ments. In addition, even if data were available for one estuary, differences like- ly exist among different recipient ecosystems and would need to be taken into account. These problems with estimating the natural invasion rate interfere with the potential to obtain a meaningful organism discharge standard that would be equivalent in its effect for all locations, or that would indeed be a desirable tar- get. There is an assumption within the NIR approach that an acceptable inva- sion rate is less than the natural invasion rate. Such assumptions make it clear that NIR is based on expert opinion as much as the IMO standards and the zero- detectable standard of CA. CONCLUSIONS The data available from which to derive scientifically based standards for living organisms in ballast discharge have been incomplete and insufficient. This situation results partly from the degree of variability possible across ballast discharges, the difficulty in garnering representative samples of live organisms in ballast discharge and the lack of a systematic approach for doing so, and the lack of baseline measures to detect the presence of newly established popula- tions of nonindigenous organisms in U.S. waters. A discharge standard based on expert opinion has provided a starting point for the regulatory process and can facilitate progression to a scientific model. For example, as ships attempt to meet the IMO D-2 standard, much of the variation in potential densities of live organisms in ballast discharge will be modulated. Advancing scientific under- standing of the probability of invasion associated with such discharges, and the 

120  Propagule Pressure and Invasion Risk in Ballast Water    performance of existing standards to reduce invasion outcomes, requires quan- titative measures of both the discharge and also associated invasions. While compliance testing provides a potential framework to achieve part of this goal, the broader integrative approach required has yet to develop (see Chapter 6). In the area of discharge limits for live organisms in ballast water, expert opinion has been tapped as an alternative to scientifically derived standard set- ting processes in light of scientific uncertainties. While each expert opinion process discussed above has conceptual merit, each is compromised by assump- tions, data limitations, or operational difficulties. Despite these drawbacks, ex- pert opinion has resulted in various standards (e.g., IMO D-2) that could provide a manageable baseline for developing scientific models that can be used to quan- titatively determine ballast water discharge standards. REFERENCES Blaschko, M. B., G. Holness, M. A. Mattar, D. Lisin, P. E. Utgoff, A. R. Hanson, H. Schultz, E. M. Riseman, M. E. Sieracki, W. M. Balch, and B. Tupper. 2005. Auto- matic in situ identification of plankton. Proceedings of the Seventh IEEE Workshop on Applications of Computer Vision 79–86. doi 10.1109/ACVMOT.2005.29. Burgman, M. 2005. Risks and decisions for conservation and environmental manage- ment. Cambridge, UK: Cambridge University Press. California State Lands Commission. 2010. 2010 Assessment of the efficacy, availabili- ty, and environmental impacts of ballast water treatment systems for use in Califor- nia waters. Sacramento, CA: California State Lands Commission. Cohen, A. N. 2005. Memo on a natural invasion rate standard. Appendix 5 In: Califor- nia State Lands Commission Report on Performance Standards for Ballast Water Discharges in California Waters. Report produced for California State Legislature. M. Falkner, L. Takata, and S. Gilmore, editors. Cohen, A. N. 2010. The natural invasion rate model and California’s ballast water dis- charge standards. Presentation to the National Academy of Science/National Rese- arch Council Committee on Assessing Numeric Limits for Living Organisms in Ba- llast Water Washington, DC. June 2, 2010. Daehler, C. C., J. S. Denslow, S. Aansari, and H.-C. Kuoa. 2004. Risk-assessment sys- tem for screening out invasive pest plants from Hawaii and other Pacific Islands. Conservation Biology 18:360–368. European Environment Agency. 2010. Invasive alien species in Europe – Assessment published May 2010. Available online at: http://www.eea.europa.eu/data-and- maps/indicators/invasive-alien-species-in-europe/invasive-alien-species-in-europe. Last accessed June 2, 2011. Faulkner, M. 2010. Presentation to the National Academy of Science/National Research Council Committee on Assessing Numeric Limits for Living Organisms in Ballast Water. June 4, 2010. Washington, DC. Gollasch, S. 2005. IMO BWMC: History, present status and challenges to the research society. Presentation to the Ballast water introductions of alien species into the Bal- tic Sea, Palanga. 

Other Approaches to Setting a Ballast Water Discharge Standard   121    International Council for the Exploration of the Sea. 2003. Comments on draft Regula- tion E-2. Submitted to the Marine Environment Protection Committee, MEPC 49/2/21. IMO (International Maritime Organization). 2011. Available online at http://www.imo. org/OurWork/Environment/BallastWaterManagement/Pages/Default.aspx#5. Last accessed June 2, 2011. IUCN (International Union for the Conservation of Nature). 2000. 100 of the World’s Worst Invasive Alien Species A selection from the Global Invasive Species Databa- se. Available online at: http://www.issg.org/booklet.pdf. Last accessed June 2, 2011. Jerde, C., A. R. Mahon, W. L. Chadderton, and D. M. Lodge. 2011. 'Sight-unseen' de- tection of rare aquatic species using environmental DNA. Conservation Letters doi: 10.1111/j.1755-263X.2010.00158. Kolar, C. S., and D. M. Lodge. 2002. Ecological predictions and risk assessment for alien fishes in North America. Science 298:1233–1236. Lee II, H., D. A. Reusser, M. Frazier, and G. Ruiz. 2010. Density Matters: Review of Approaches to Setting Organism-Based Ballast Water Discharge Standards. EPA/600/R-10/031. EPA Office of Research and Development, National Health and Environmental Effects Research Laboratory, Western Ecology Division. Maguire, L. A. 2004. What can decision analysis do for invasive species management? Risk 24:859–868. Mahon, A. R., M. Barnes, S. Senapati, J. Darling, H.-C. Chang, J. L. Feder, and D. M. Lodge. 2011. Molecular detection of invasive species in heterogeneous mixtures using a microfluidic carbon platform. PLoS One 6(2):e17280. Miller, A. W., M. Frazier, G. E. Smith, E. S. Perry, G. M. Ruiz, and M. N. Tamburri. 2011. Enumerating sparse organisms in ships' ballast water: why counting to 10 is not so easy. Environmental Science and Technology (in press). dx.doi.org/ 10.1021/es102790d. Minton, M., E. Verling, A. W. Miller, and G. M. Ruiz. 2005. Reducing propagule sup- ply by ships to limit coastal invasions: effects of emerging strategies. Frontiers in Ecology and the Environment 6:304–308. Paini, D. R., S. P. Worner, D. C. Cook, P. J. DeBarro, and M. B. Thomas. 2010. Using a self-organizing map to predict invasive species: sensitivity to data errors and a com- parison with expert opinion. Journal of Applied Ecology 47:290–298. Phelong, P. C., P. A. William, and S. R. Halloy. 1999. A weed risk assessment model for use as a biosecurity tool evaluating plant introductions. Journal of Environmen- tal Management 57:239–251. Roy, K., D. Jablonski, and J. W. Valentine. 1995. Thermally anomalous assemblages revisited: patterns in the extraprovincial range shifts of Pleistocene marine mollusks. Geology 23:1071–1074. Tucker, K. C., and D. M. Richardson. 1995. An expert system for screening potentially invasive alien plants in South African Fynbos. Journal of Environmental Manage- ment 44:309–338. Vermeij, G. J. 1991. When biotas meet: understanding biotic interchange. Science 253: 1099–1104.